As a quick follow-up to Meredith's post, I wanted to add a few additional thoughts.
Meredith mentioned that "at IPA, we always strive to conduct research in a way that identifies not just whether an isolated solution works, but why it works, so that we gain information about what was causing the problem in the first place.' The majority of IPA's projects involve randomized evaluations in the developing world, and we have been able to replicate similar evaluations in various contexts in order to understand the local factors that play a role in a program's impact. While individual randomized evaluations may be criticized for their external validity, the replication approach helps to address that concern. Ultimately, numerous points of light shine through and we begin to develop a holistic perspective on things.
Which is not to say that randomized evaluations are the only way to go about things; but they play an integral part in the development of functional problem-solving systems. This is something that top-down development advocates would do well to take note of. For instance, development theorists are constantly on the lookout for the "right" institutions. As Banerjee (2008) observes, the institutionalist literature is still unclear as to what sort of institutions need to be encouraged, with few reliable policy prescriptions having emerged. If you subscribe to the thinking of seminal institutionalist thinker Douglass North, institutional change is overwhelmingly an incremental one. Alternatively, you could subscribe to the original Shock Doctrine that advocated inter alia rapid, wholesale change.
Although I don't quite subscribe to all the rhetoric of shock doctrine critics, I think there is enough reason to believe that institutional change should be an incremental process; after all, when we barely know what institutions truly work, how could we advocate system-wide changes? Identifying the right institutions in context-specific programs (via randomized evaluations, for example) and gradually working one's way up seems like a better manifesto for success.
Finally, if you want to hear more about Bill Easterly's thoughts on how to approach development, listen to a podcast that he did with George Mason's Russ Roberts in February 2008, on Roberts' weekly podcast EconTalk. The podcast is a little dated, but the topics remains ever-relevant and of interest. It makes for a great listen and contains, amongst other things, a comparison of the relative merits of the sort of approach advocated by Easterly versus the top-down approach favored by economists like Jeffrey Sachs.
BANERJEE, A., (2008), "Big Answers for big questions: the presumption of growth policy," Unpublished Transcript of Speech Prepared for the Brookings Conference on "What Works in Development? Thinking Big and Thinking Small".
NORTH, D., (1990), Institutions, Institutional Change and Economic Performance (Cambridge: Cambridge University Press, 1990).